Neuroscience methods

Evaluating methods

What are we measuring?

  • Structure
  • Activity
    • Why not function?

What is the question?

  • Structure X -> Structure Y
  • Structure X -> Function Y

Evaluating methods

Strengths & Weaknesses

  • Cost
  • Invasiveness
  • Spatial/temporal resolution

…and temporal resolution

(Sejnowski, Churchland, & Movshon, 2014)

Types of methods

  • Structural
    • Anatomy
    • Connectivity/connectome
  • Functional
    • What does it do?
    • Physiology/Activity

Structural methods

Mapping microstructure

  • Cell/axon stains
  • Cellular distribution, concentration, microanatomy

Golgi stain

  • whole cells, but small %

http://connectomethebook.com/wp-content/uploads/2011/11/Brainforest17_1119.jpg

http://wam.umn.edu/wp-content/uploads/2016/12/WAM_Cajal_m1673.jpg

http://wam.umn.edu/wp-content/uploads/2016/12/WAM_Cajal_m1673.jpg

Nissl stain

  • Only cell bodies
  • Cell density ~ color intensity

CUBIC

  • CUBIC (“clear, unobstructed brain/body imaging cocktails and computational analysis”)
  • (Susaki et al., 2014)

Example

[[@Ando2018-ze]](https://doi.org/10.1016/B978-0-444-63639-3.00021-9)

(Ando, Laborde, Brion, & Duyckaerts, 2018)

Fig. 21.2. Immunostaining of CUBIC-clarified 500-μm-thick slices from human Alzheimer disease postmortem brain frontal cortex. Human Alzheimer disease frontal cortex tissue immunostained for Aß (6E10, red) and for tau (B19, green). Stack depth of 264 μm; step size = 1 μm. Stack photos were taken with a two-photon microscope equipped with a 20 × air objective. Scale bar, 100 μm.

Evaluating micro/cellular techniques

  • Invasive (in humans post-mortem only)
  • High spatial resolution, but poor/coarse temporal

Example

“SHANK2 mutations associated with autism spectrum disorder cause hyperconnectivity of human neurons” (Zaslavsky et al., 2019)

[[@Zaslavsky2019-ln]](http://dx.doi.org/10.1038/s41593-019-0365-8)

(Zaslavsky et al., 2019)

a, iPSCs generated from multiple control and affected individuals are differentiated into NPCs. NPCs are differentiated in separate wells for 4 weeks and then differentially fluorescently labeled control (CTRL) and mutant (MUT) cells are sparsely seeded onto a large unlabeled neuronal population (the lawn) and cocultured with astrocytes. b, Timeline of the experiment, starting with seeding of NPCs. Measurements of mutant cells are normalized to control cells in the same well. c, Sparse seeding allows simultaneous analyses of cell morphology and connectivity (total number of SYN1 puncta) of single neurons. Scale bars, 100 μm. d, To compare cell morphology, paired representative traces are shown of control and SHANK2 ASD or engineered SHANK2 KO neurons grown in the same well. e, To compare synaptic function, sEPSCs are recorded from neurons grown in the same well. Confocal images and traces shown in c and d are representative of iPSC-derived neurons imaged in experiments depicted in Fig. 2a–c. sEPSC traces shown in e are representative of patch-clamp recordings of iPSC-derived neurons described in Fig. 3.

Mapping macro-structures

Magnetic Resonance Imaging (MRI)

What it measures/how it works

  • Magnetic resonance a property of some isotopes and complex molecules
  • Hydrogen (\(H\)), common in water & fat, is one
  • In magnetic field, \(H\) atoms absorb and release radio frequency (RF) energy
  • \(H\) atoms align with strong magnetic field
  • Applying RF pulse perturbs alignment
  • Rate/timing of realignment varies by tissue
  • Realignment gives off radio frequency (RF) signals
  • Strength of RF ~ density of \(H\) (or other target)
  • K-space (frequency/phase) -> anatomical space

http://s.hswstatic.com/gif/mri-steps.jpg

Structural MRI

  • Tissue density/type differences
  • Gray matter (nerve cells & dendrites) vs. white matter (axon fibers)

MR Spectroscopy (specific metabolites)

  • Region sizes/volumes

Voxel-based morphometry (VBM)

  • MRI technique for measuring brain sizes/volumes
[[@Pomarol-Clotet2010-tq]](https://dx.doi.org/10.1038/mp.2009.146)

(Pomarol-Clotet et al., 2010)

  • Volume differences in schizophrenics vs. controls
  • Colored portions are statistical maps placed on top of a base structural map. - Maps provide information about the comparison in brain volumes between patients and controls in those areas

Mapping the wiring diagram (“connectome”)

Retrograde (output -> input) vs. anterograde (input -> output) tracers

http://openi.nlm.nih.gov/imgs/512/348/3176268/3176268_1471-2105-12-351-2.png

Diffusion Tensor Imaging (DTI)

  • Structural MRI technique
  • Diffusion tensor: measurement of spatial pattern of \(H_2O\) diffusion in small volume
  • Uniform (“isotropic”) vs. non-uniform (“anisotropic”)
  • Strong anisotropy suggests large # of axons with similar orientations (fiber tracts)

Visualizing the connectome

Functional methods

  • Recording from the brain
  • Interfering with the brain
  • Stimulating the brain
  • Simulating the brain

Recording from the brain

Single/multi-unit Recording

  • Microelectrodes + amplification
  • Small numbers of nerve cells
https://www.nature.com/nrn/journal/v5/n11/images/nrn1535-i1.jpg

https://www.nature.com/nrn/journal/v5/n11/images/nrn1535-i1.jpg

  • What does neuron X respond to?
  • How does firing frequency, timing vary with behavior?
  • Great temporal (ms), spatial resolution (um)
  • Invasive
  • Rarely suitable for humans, but…

Electrocorticography (ECoG)

Grid electrodes: (A) Craniotomy performed for electrocorticography (ECoG) grid electrode placement in epilepsy surgery candidate at Comprehensive Epilepsy Program, Florida Hospital for Children, Orlando, Florida, United States. (B) ECoG electrode grids placed directly on the brain surface. They will be used during presurgical monitoring for localizing seizure onset zone. The same electrodes are stimulated during electrical cortical stimulation mapping for identification of eloquent cortex. The ECoG signal recorded from these grids is separated in a different stream and used for real-time functional mapping (RTFM). (C) 3D reconstruction of the brain with overlaid grid electrodes. This reconstruction is used for creating RTFM montage.

Story about child who underwent ECoG surgery.

Positron Emission Tomography (PET)

  • Radioactive tracers (glucose, oxygen)
  • Positron decay activates paired detectors
  • Tomographic techniques reconstruct 3D geometry
  • Experimental condition - control
  • Average across individuals
  • Temporal (~ s) and spatial (mm-cm) resolution worse than fMRI
  • Radioactive exposures + mildly invasive
  • Dose < airline crew exposure in 1 yr

Functional Magnetic Resonance Imaging (fMRI)

  • Neural activity -> local \(O_2\) consumption increase
  • Blood Oxygen Level Dependent (BOLD) response
  • Oxygenated vs. deoxygenated hemoglobin ≠ magnetic susceptibility
  • How do regional blood \(O_2\) levels (& flow & volume) vary with behavior X?
  • MRI “signals” relate to the speed (1/T) of “relaxation” of the perturbed nuclei to their state of alignment with the main (\(B_0\)) magnetic field.
  • Imaging protocols emphasize different time constants of this relaxation (\(T1\), \(T2\), \(T2^*\)); \(T^2*\) for BOLD imaging

Evaluating fMRI

  • Non-invasive, but expensive
  • Moderate but improving (mm) spatial, temporal (~sec) resolution
  • Spatial limits due to
    • field strength (@ 3T \(~3mm^3\) voxel)
    • Physiology of hemodynamic response
  • Temporal limits due to
    • Hemodynamic Response Function (HRF): ~ 1s delay plus 3-6 s ramp-up
    • Speed of image acquisition
  • Indirect measure of neural activity

Hemodynamic Response Function (HRF)

https://openi.nlm.nih.gov/imgs/512/236/3109590/3109590_TONIJ-5-24_F1.png

https://openi.nlm.nih.gov/imgs/512/236/3109590/3109590_TONIJ-5-24_F1.png

Generate “predicted” BOLD response to event; compare to actual

Higher field strengths (3 Tesla vs. 7 Tesla)

[[@Sladky2013-bu]](https://dx.doi.org/10.1016/j.ejrad.2011.09.025)

(Sladky et al., 2013)

but fMRI underpowered

(Szucs & Ioannides, 2017)

“Assuming a realistic range of prior probabilities for null hypotheses, false report probability is likely to exceed 50% for the whole literature.”

(Szucs & Ioannides, 2017)

Functional Near-infrared Spectroscopy (fNIRS)

  • Near infrared light penetrates scalp and skull, refracted by brain tissue
  • Returned signal altered by blood \(O_2\) levels
  • Time course (temporal resolution) ~ BOLD fMRI
  • Spatial resolution low
  • More suitable for pediatric populations (less susceptible to movement artefact)
Source: https://cibsr.stanford.edu/NIRS_Lab.html

Source: https://cibsr.stanford.edu/NIRS_Lab.html

Source: https://nirx.net

Source: https://nirx.net

Electroencephalography (EEG)

  • How does it work?
  • Electrodes on scalp or brain surface

What does EEG measure?

  • Voltage differences between source and reference electrode
  • Combined activity of huge # of neurons
  • Current/voltage gradients between apical (near surface) dendrites and basal (deeper) dendrites and cell body/soma
https://neurofeedbackalliance.org/wp-content/uploads/2016/10/Dipole.jpg"

https://neurofeedbackalliance.org/wp-content/uploads/2016/10/Dipole.jpg"

Evaluating EEG

  • High temporal, poor spatial resolution
  • Analyze activity in different ‘bands’ of frequencies
    • LOW: deep sleep (delta or \(\delta\) band)
    • MIDDLE: Quiet, alert state (alpha \(\alpha\) band)
    • HIGHER: Sensorimotor activity reflecting observed actions? (mu or \(\mu\) band), (Hobson & Bishop, 2017)
    • HIGHER STILL: “Binding” information across senses or plasticity? (gamma or \(\gamma\) band), (Amo et al., 2017)
https://www.peakmind.co.uk/images/frequency.jpg

https://www.peakmind.co.uk/images/frequency.jpg

Brain Computer Interface (BCI)

  • Based on EEG/ERPs

Magneto-encephalography (MEG)

  • Like EEG, but measuring magnetic fields
  • Electrical and magnetic fields orthogonal
  • High temporal resolution
  • Magnetic fields propagate w/o distortion
    • But are orthogonal to electric field
  • Requires shielded chamber (to keep out strong magnetic fields)
  • ++ cost vs. EEG
https://upload.wikimedia.org/wikipedia/commons/e/e6/NIMH_MEG.jpg

https://upload.wikimedia.org/wikipedia/commons/e/e6/NIMH_MEG.jpg

New device minimizes problems with motion
[[@Hill2019-ik]](https://doi.org/10.1038/s41467-019-12486-x)

(Hill et al., 2019)

Figure 1. A paediatric MEG system: a Experimental setup for three participants age 2- (left), 5- (centre) and 24-years (right). OPMs, housed in a modified bike helmet, measured the MEG signal. b Time-frequency spectra from a single (synthesised gradiometer) channel. Changes in neural oscillations are shown; blue indicates a reduction in oscillatory amplitude relative to baseline; yellow indicates an increase. Note reduction in beta (13–30 Hz) and mu (8–13 Hz) amplitude. c The spatial signature of beta modulation during the period of tactile stimulation (0 s < t < 2 s) (blue overlay)

How do EEG/MEG and fMRI relate?

(Logothetis, Pauls, Augath, Trinath, & Oeltermann, 2001)

Manipulating the brain

  • Interfering with it
  • Stimulating it

Interfering with the brain

  • Nature’s“experiments”
  • Stroke, head injury, tumor
  • Neuropsychology

Phineas Gage

Evaluating neuropsychological methods

  • Logic: IF damage to area X impairs performance, THEN region critical for behavior Y
  • Double dissociation: Damage to area Z leaves behavior Y intact
  • Weak spatial/temporal resolution

Stimulating the brain

  • Electrical (Direct Current Stimulation - DCS)
  • Pharmacological
  • Magnetic (Transcranial magnetic stimulation - TMS)
  • Spatial/temporal resolution?
  • Assume stimulation mimics natural activity?

Optogenetics

  • Gene splicing techniques insert light-sensitive molecules into neuronal membranes
  • Application of light at specific wavelengths alters neuronal function
  • Cell-type specific and temporally precise control
  • Mimics brain activity

https://youtu.be/FlGbznBmx8M

Simulating the brain

  • Computer/mathematical models of brain function
  • Example: neural networks
  • Cheap, noninvasive, can be stimulated or “lesioned”

Blue Brain project

Markram, 2006

References

Amo, C., De Santiago, L., Zarza Luciáñez, D., León Alonso-Cortés, J. M., Alonso-Alonso, M., Barea, R., & Boquete, L. (2017). Induced gamma band activity from EEG as a possible index of training-related brain plasticity in motor tasks. PloS One, 12(10), e0186008. https://doi.org/10.1371/journal.pone.0186008
Ando, K., Laborde, Q., Brion, J.-P., & Duyckaerts, C. (2018). Chapter 21 - 3D imaging in the postmortem human brain with CLARITY and CUBIC. In I. Huitinga & M. J. Webster (Eds.), Handbook of clinical neurology (Vol. 150, pp. 303–317). Elsevier. https://doi.org/10.1016/B978-0-444-63639-3.00021-9
Hill, R. M., Boto, E., Holmes, N., Hartley, C., Seedat, Z. A., Leggett, J., … Brookes, M. J. (2019). A tool for functional brain imaging with lifespan compliance. Nature Communications, 10(1), 4785. https://doi.org/10.1038/s41467-019-12486-x
Hobson, H. M., & Bishop, D. V. M. (2017). The interpretation of mu suppression as an index of mirror neuron activity: Past, present and future. Royal Society Open Science, 4(3), 160662. https://doi.org/10.1098/rsos.160662
Lee, J. K., Andrews, D. S., Ozonoff, S., Solomon, M., Rogers, S., Amaral, D. G., & Nordahl, C. W. (2021). Longitudinal evaluation of cerebral growth across childhood in boys and girls with autism spectrum disorder. Biological Psychiatry, 90(5), 286–294. https://doi.org/10.1016/j.biopsych.2020.10.014
Lichtman, J. W., Livet, J., & Sanes, J. R. (2008). A technicolour approach to the connectome. Nature Reviews Neuroscience, 9(6), 417–422. https://doi.org/10.1038/nrn2391
Logothetis, N. K., Pauls, J., Augath, M., Trinath, T., & Oeltermann, A. (2001). Neurophysiological investigation of the basis of the fMRI signal. Nature, 412(6843), 150–157. https://doi.org/10.1038/35084005
Logothetis, N. K., & Wandell, B. A. (2004). Interpreting the BOLD signal. Annu. Rev. Physiol., 66(1), 735–769. https://doi.org/10.1146/annurev.physiol.66.082602.092845
Pomarol-Clotet, E., Canales-Rodrı́guez, E. J., Salvador, R., Sarró, S., Gomar, J. J., Vila, F., … McKenna, P. J. (2010). Medial prefrontal cortex pathology in schizophrenia as revealed by convergent findings from multimodal imaging. Mol. Psychiatry, 15(8), 823–830. https://doi.org/10.1038/mp.2009.146
Sejnowski, T. J., Churchland, P. S., & Movshon, J. A. (2014). Putting big data to good use in neuroscience. Nat. Neurosci., 17(11), 1440–1441. https://doi.org/10.1038/nn.3839
Sladky, R., Baldinger, P., Kranz, G. S., Tröstl, J., Höflich, A., Lanzenberger, R., … Windischberger, C. (2013). High-resolution functional MRI of the human amygdala at 7 T. Eur. J. Radiol., 82(5), 728–733. https://doi.org/10.1016/j.ejrad.2011.09.025
Susaki, E. A., Tainaka, K., Perrin, D., Kishino, F., Tawara, T., Watanabe, T. M., … Ueda, H. R. (2014). Whole-brain imaging with single-cell resolution using chemical cocktails and computational analysis. Cell, 157(3), 726–739. https://doi.org/10.1016/j.cell.2014.03.042
Zaslavsky, K., Zhang, W.-B., McCready, F. P., Rodrigues, D. C., Deneault, E., Loo, C., … Ellis, J. (2019). SHANK2 mutations associated with autism spectrum disorder cause hyperconnectivity of human neurons. Nature Neuroscience, 22(4), 556–564. https://doi.org/10.1038/s41593-019-0365-8